INDEX

Explanations

phrases related to technical instructions and settings

oai_token-act-pair · gpt-3.5-turbo

technical instructions and settings configurations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wat

-0.76

 republican

-0.71

atton

-0.67

onna

-0.66

 retard

-0.66

eno

-0.66

 upset

-0.65

 loyal

-0.65

 transpl

-0.64

enos

-0.64

POSITIVE LOGITS

 âĩ

0.88

below

0.85

 appendix

0.85

tips

0.83

 section

0.83

 Preferences

0.83

 below

0.82

 Settings

0.80

section

0.79

sections

0.79

Activations Density 0.432%

phrases related to technical instructions and settings

technical instructions and settings configurations

No Comments

No Known Activations